After upgrade from 7.0.5 to 7.1.4 on Raspbian Stretch seahub fails to start

framp · June 22, 2020, 1:43pm

Today I upgraded from 7.0.5 to 7.1.4 on Stretch. I followed https://download.seafile.com/published/seafile-manual/upgrade/upgrade_notes_for_7.1.x.md. seahub doesn’t start any more :-(. I found Seahub fails after upgrade to 7.1.4 but this one talks about upgrade on buster and talks about a module not found error message. But I frankly don’t get any error message

sudo  systenctl start seahub
Job for seahub.service failed because the control process exited with error code. 
See "systemctl status seahub.service" and "journalctl -xe" for details.
 journalctl -xe -u seahub
 -- Subject: Unit seahub.service has begun start-up
    -- Defined-By: systemd
-- Support: https://www.debian.org/support
-- 
-- Unit seahub.service has begun starting up.
Jun 22 15:21:49 seafile seahub.sh[999]: LC_ALL is not set in ENV, set to en_US.UTF-8
Jun 22 15:21:49 seafile seahub.sh[999]: Starting seahub at port 8000 ...
Jun 22 15:22:06 seafile systemd[1]: Started Seafile hub.
-- Subject: Unit seahub.service has finished start-up
-- Defined-By: systemd
-- Support: https://www.debian.org/support
-- 
-- Unit seahub.service has finished starting up.
-- 
-- The start-up result is done.
Jun 22 15:22:07 seafile systemd[1]: seahub.service: Main process exited, code=exited, status=1/FAILURE
Jun 22 15:22:07 seafile systemd[1]: seahub.service: Unit entered failed state.
Jun 22 15:22:07 seafile systemd[1]: seahub.service: Failed with result 'exit-code'.
Jun 22 15:32:08 seafile systemd[1]: Starting Seafile hub...
-- Subject: Unit seahub.service has begun start-up
-- Defined-By: systemd
-- Support: https://www.debian.org/support
-- 
-- Unit seahub.service has begun starting up.
Jun 22 15:32:08 seafile seahub.sh[1605]: LC_ALL is not set in ENV, set to en_US.UTF-8
Jun 22 15:32:08 seafile seahub.sh[1605]: Starting seahub at port 8000 ...
Jun 22 15:32:14 seafile systemd[1]: seahub.service: Control process exited, code=exited status=1
Jun 22 15:32:14 seafile systemd[1]: Failed to start Seafile hub.
-- Subject: Unit seahub.service has failed
-- Defined-By: systemd
-- Support: https://www.debian.org/support
-- Unit seahub.service has failed.

– The result is failed.
Jun 22 15:32:14 seafile systemd[1]: seahub.service: Unit entered failed state.
Jun 22 15:32:14 seafile systemd[1]: seahub.service: Failed with result ‘exit-code’.

seahub.log is empty.

When I invoke bash -x ./seahub.sh the last lines I get are

+ python3 /seafile-data/home/buddy/seafile/seafile-server-7.1.4/check_init_admin.py
+ python3 /seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/bin/gunicorn seahub.wsgi:application -c /seafile-data/home/buddy/seafile/conf/gunicorn.conf.py --preload
+ sleep 5
+ pgrep -f seahub.wsgi:application
+ printf '\033[33mError:Seahub failed to start.\033[m\n'
Error:Seahub failed to start.
+ echo 'Please try to run "./seahub.sh start" again'
Please try to run "./seahub.sh start" again
+ exit 1

Would be great somebody knows how can I can get more information about the root cause of gunicorn to fail

cat gunicorn.conf.py
import os

daemon = True
workers = 5

bind = “127.0.0.1:8000”

pids_dir = ‘/seafile-data/home/buddy/seafile/pids’
pidfile = os.path.join(pids_dir, ‘seahub.pid’)

timeout = 1200

limit_request_line = 8190

Just ask for any other details/configs which will help to find the root cause of the miracle

UPDATE:
When I execute the following command used to start gunicorn I get

python3 /seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/bin/gunicorn seahub.wsgi:application -c /seafile-data/home/buddy/seafile/conf/gunicorn.c
onf.py --preload
Traceback (most recent call last):
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/bin/gunicorn”, line 5, in
from gunicorn.app.wsgiapp import run
ImportError: No module named ‘gunicorn’

May this be the root cause of the issue?

axstet · June 22, 2020, 5:55pm

Running Seahub in FastCGI mode should give you better (more detailed) error messages.

framp · June 22, 2020, 6:10pm

Thank you very much for this hint. That’s what I get now:

./seahub.sh start-fastcgi 8000

Starting seahub (fastcgi) at 127.0.0.1:8000 …
Traceback (most recent call last):
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/seahub/avatar/models.py”, line 23, in
from PIL import Image
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/PIL/Image.py”, line 93, in
from . import _imaging as core
ImportError: cannot import name ‘_imaging’

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/manage.py”, line 10, in
execute_from_command_line(sys.argv)
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/django/core/management/init.py”, line 364, in execute_from_command_line
utility.execute()
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/django/core/management/init.py”, line 338, in execute
django.setup()
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/django/init.py”, line 27, in setup
apps.populate(settings.INSTALLED_APPS)
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/django/apps/registry.py”, line 108, in populate
app_config.import_models()
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/thirdpart/django/apps/config.py”, line 202, in import_models
self.models_module = import_module(models_module_name)
File “/usr/lib/python3.5/importlib/init.py”, line 126, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File “”, line 986, in _gcd_import
File “”, line 969, in _find_and_load
File “”, line 958, in _find_and_load_unlocked
File “”, line 673, in _load_unlocked
File “”, line 673, in exec_module
File “”, line 222, in _call_with_frames_removed
File “/seafile-data/home/buddy/seafile/seafile-server-7.1.4/seahub/seahub/avatar/models.py”, line 26, in
import Image
ImportError: No module named ‘Image’
Error:Seahub failed to start.

axstet · June 22, 2020, 6:26pm

You need python 3.6 or newer. Please read all of:

framp · June 22, 2020, 6:43pm

Thank you very much for the link.

You are right. I run right now

python3 --version
Python 3.5.3

I followed this page which unfortunately doesn’t refer to your link. I will upgrade my python now.

shoeper · June 22, 2020, 7:02pm

Does it still work?

framp · June 22, 2020, 7:07pm

No. seahub doesn’t start any more. I just followed these instructions

When I execute ```
pip3 install --timeout=3600 Pillow pylibmc captcha jinja2 sqlalchemy \ django-pylibmc django-simple-captcha python3-ldap


I get a lot of error messages:

>  running bdist_wheel
>   running build
>   running build_py
>   creating build
>   creating build/lib.linux-armv7l-3.5
>   creating build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/consts.py -> build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/__init__.py -> build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/autoconf.py -> build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/__main__.py -> build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/test.py -> build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/pools.py -> build/lib.linux-armv7l-3.5/pylibmc
>   copying src/pylibmc/client.py -> build/lib.linux-armv7l-3.5/pylibmc
>   running build_ext
>   building '_pylibmc' extension
>   creating build/temp.linux-armv7l-3.5
>   creating build/temp.linux-armv7l-3.5/src
>   arm-linux-gnueabihf-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -g -fdebug-prefix-map=/build/python3.5-6waWnr/python3.5-3.5.3=. -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -DUSE_ZLIB -I/usr/include/python3.5m -c src/_pylibmcmodule.c -o build/temp.linux-armv7l-3.5/src/_pylibmcmodule.o -fno-strict-aliasing -std=c99
>   In file included from src/_pylibmcmodule.c:34:0:
>   src/_pylibmcmodule.h:42:36: fatal error: libmemcached/memcached.h: No such file or directory
>    #include <libmemcached/memcached.h>
>                                       ^
>   compilation terminated.
>   error: command 'arm-linux-gnueabihf-gcc' failed with exit status 1
>   
>   ----------------------------------------
>   Failed building wheel for pylibmc
>   Running setup.py clean for pylibmc
> Failed to build pylibmc
> Installing collected packages: Pillow, pylibmc, captcha, MarkupSafe, jinja2, sqlalchemy, django-pylibmc, six, sqlparse, pytz, Django, django-ranged-response, django-simple-captcha, pyasn1, python3-ldap
>   Running setup.py install for pylibmc ... error
>     Complete output from command /usr/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-wonnoae8/pylibmc/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-wii__cum-record/install-record.txt --single-version-externally-managed --compile --user --prefix=:
>     running install
>     running build
>     running build_py
>     creating build
>     creating build/lib.linux-armv7l-3.5
>     creating build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/consts.py -> build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/__init__.py -> build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/autoconf.py -> build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/__main__.py -> build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/test.py -> build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/pools.py -> build/lib.linux-armv7l-3.5/pylibmc
>     copying src/pylibmc/client.py -> build/lib.linux-armv7l-3.5/pylibmc
>     running build_ext
>     building '_pylibmc' extension
>     creating build/temp.linux-armv7l-3.5
>     creating build/temp.linux-armv7l-3.5/src
>     arm-linux-gnueabihf-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -g -fdebug-prefix-map=/build/python3.5-6waWnr/python3.5-3.5.3=. -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -DUSE_ZLIB -I/usr/include/python3.5m -c src/_pylibmcmodule.c -o build/temp.linux-armv7l-3.5/src/_pylibmcmodule.o -fno-strict-aliasing -std=c99
>     In file included from src/_pylibmcmodule.c:34:0:
>     src/_pylibmcmodule.h:42:36: fatal error: libmemcached/memcached.h: No such file or directory
>      #include <libmemcached/memcached.h>
>                                         ^
>     compilation terminated.
>     error: command 'arm-linux-gnueabihf-gcc' failed with exit status 1
>     
>     ----------------------------------------
> Command "/usr/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-wonnoae8/pylibmc/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-wii__cum-record/install-record.txt --single-version-externally-managed --compile --user --prefix=" failed with error code 1 in /tmp/pip-build-wonnoae8/pylibmc/

:-(

axstet · June 22, 2020, 7:30pm

Not sure for production, but it works fine to get detailed python errors in case seahub.sh does not start.

framp · June 22, 2020, 8:54pm

Not sure I get you. Do you say even these error messages pop up seahub will work? I use this seafile server in production and want to have a clean environment. Actually the upgrade of python already fails and thus seahub still doesn’t start up.

Reading the doc about the upgrade to 7.1.4 it looks like there has a lot of stuff changed - not only the python upgrade. It’s even worse seafdav has issues in 7.1.4 because I need this functionality.

Does it make sense to setup buster and install 7.1.4 on top because the buster version of 7.1.4 is more reliable than the stretch version? Or should I consider 7.1.4 as a buggy release and revert to 7.0.5?

squirrel · June 22, 2020, 10:30pm

I only see in your latest error message that libmemcached/memcached.h is not found.
I resolved this error (on Ubuntu 20.04, but still) by installing libmemcached-dev - which seems to be available for stretch.

Yet, I don’t know how much this will help you as your python version is still 3.5 which does not seem to work anyway. (Again, no experience myself, just what I’ve read.)

elghoto · June 23, 2020, 5:48am

From a fresh install I noticed that even if you update to python3.6+ (in my case python3.7) you have issues with finding packages. You will need to link seafile-server-7.1.4/seafile/lib/python3.7 -> seafile-server-7.1.4/seafile/lib/python3.6 and that makes thing to work.

axstet · June 23, 2020, 5:30pm

Correct. It’s all documented in the referenced post.

framp · June 23, 2020, 5:56pm

I had serious issues to upgrade python on stretch to 7. So I decided to setup a new system with buster. Again I had serious issues to get seafile 7.1.4 up and running. I now revert to 7.0.5 to get my seafile up and running again. I have to have a running seafile again and cannot spend any more time on trial and error install of 7.1.4.

I run seafile for a long time already and upgraded seafile a lot of times without any issues on my production system successfully. When wsgi was introduced I decided to setup a test system to check how to upgrade and how the upgrade works and then started the upgrade on my production system. Looks like I should have done this again when upgrading from 7.0.5 to 7.1.4. I hope next upgrade to 7.1.5 will run again smoothly and there is no need to setup a dedicated test environment first

axstet · June 23, 2020, 7:40pm

Seems you were having trouble switching to python 3. Since 7.1.5 will by python 3 as well, I assume you will have the same issues all over again. Worth looking into your problems when you got time.

framp · June 23, 2020, 8:29pm

buster already has python 7. I frankly just stopped when I got seahub to fail to start on buster and didn’t spend any time to isolate the root cause because I was frustated. There was no need for me to upgrade to 7.1.4 but given my past experiences with seafile upgrades I thought it’s worth to upgrade to 7.1.4 … but looks like this was a bad idea.

For now I will stick with 7.0.5 which works fine for me. Maybe I will give it another try to install 7.1.4 on a test system when I have some spare time …

elghoto · June 23, 2020, 11:22pm

Once you pass the “it runs phase” there are more issues as seen in other threads. I’m happy that 7.0.5 worked for you, I’m switching back too.

framp · June 26, 2020, 7:57pm

I just got my seafile up and running with version 7.0.5 with my backups. It took me a while because I never had to use my seafile backups before to revert after an upgrade to a weak version and thus I never tested the restore before.

I’m not sure how I’m going to make sure this incident will not happen any more. For now I will stick with 7.0.5 until Stretch has EOS. Hope then there will be a new version of seafile available which allows a smooth upgrade without any issues as it was in all previous upgrades.